Statistical Indications of Toponium Formation in Top Quark Pair Production

Abstract

We present an analysis of six differential cross-section measurements of top-quark pair production in the dilepton channel from the ATLAS and CMS experiments. The data are compared to state-of-the-art QCD predictions with and without the inclusion of toponium formation effects. This contribution is modelled via a re-weighting of fixed-order matrix elements using the Green's function of the non-relativistic QCD Hamiltonian, and we employ a statistical model to quantify the preference of the data for the toponium hypothesis. All observables yield Bayes factors larger than unity, with two exceeding 20, yielding strong evidence for toponium formation in top-quark pair production at the LHC.

Figures (5)

• Figure 1: Data in the two-dimensional differential cross section of $\text{m}_{\ell \ell}\xspace$ and $\Delta \phi (\ell \ell)\xspace$ATLAS:2019hau are compared with predictions excluding (blue) and including a toponium component, using either the predicted (red) or best-fit (green) cross section. The upper panel shows the data and model predictions; the middle panel shows the ratio of the expected toponium signal to the experimental uncertainty in each bin; and the lower panel shows the ratio of the data to each prediction with the shaded band representing the total uncertainty on the data.
• Figure 2: Values of the estimated toponium cross section $\hat{\sigma}_{\rm toponium}$ for each observable (dashed lines with shaded bands). The theoretical prediction from Fuks:2024yjj with a 50% uncertainty is shown as a grey band, and the ATLAS detector-level measurement ATLAS:2025kvb is shown with a green band.
• Figure 3: Same as Figure \ref{['fig:atlas_mlldelphill']} but for $\frac{\mathrm{d}\sigma}{\mathrm{d}\text{m}_{\ell \ell}\xspace}$ as measured from ATLAS:2019hau (left) and CMS:2018adi (right).
• Figure 4: Same as Figure \ref{['fig:atlas_mlldelphill']} but for $\frac{\mathrm{d}\sigma}{\mathrm{d}\Delta \phi (\ell \ell)\xspace}$ as measured from ATLAS:2019hau (left) and CMS:2018adi (right).
• Figure 5: Same as Figure \ref{['fig:atlas_mlldelphill']} but for $\frac{\mathrm{d}\sigma}{\mathrm{d}\Delta \eta (\ell \ell)\xspace}$ as measured from CMS:2018adi.